In 1995 the Phoenix Epidemiology and Clinical Research Branch of NIDDK conducted a cross-sectional study to identify the effects of traditional and western environments on prevalence of type 2 diabetes mellitus and obesity in Pima Indians in Mexico and the United States. The Mexican Pimas live in a remote mountainous region in the village of Maycoba and in 1995 had experienced little change from their traditional lifestyle. A similar number of non-Pima Mexicans live in Maycoba. In contrast to the Mexican Pimas, the U.S. Pimas live in a more westernized society and they have a high prevalence of diabetes. The survey found that although diabetes and obesity were more prevalent in Mexican Pimas than non-Pimas, both Mexican groups had a significantly lower prevalence of these disorders than U.S. Pimas. The present project is being conducted in collaboration with investigators from Northern Arizona University and Centro de Investigacion en Alimentacion y Desarollo in Hermosillo, Mexico. Another survey is being undertaken among the Pima and non-Pima residents of Maycoba, using the same methods used in the 1995 survey, to determine the extent to which the prevalence of T2DM, obesity and related conditions have changed since the original survey. The main survey will attempt to enroll as many of the residents of Maycoba as possible. In a subset of the population (100 individuals), resting and total energy expenditure are being measured by indirect calorimetry using a ventilated hood and by doubly-labeled water, respectively. From these measurements, physical activity energy expenditure will be calculated. An ethnographic survey has also been conducted in a different subset of individuals to collect information about health beliefs, environmental changes in the community and oral health practices. The surveys in Maycoba have now been completed. In the main survey a total of 657 residents of Maycoba have participated; 369 of the individuals are of at least partial Pima heritage. This constitutes 84% of the eligible Pima population of Maycoba, and 81% of the eligible non-Pima population. The prevalence of diabetes increased in non-Pima men and in both non-Pima and Pima women, but not in Pima men. Prevalence of obesity increased in all groups, and obesity was strongly associated with the number of modern technological features to which an individual had access, suggesting it is influenced by modernization. Additional analyses of diet, physical activity and of genetic polymorphisms are currently underway. Analyses of 80 established diabetes or obesity variants suggest that most established type 2 diabetes variants do not differ in prevalence between Mexican Pimas and Arizona Pimas; the frequency of established obesity variants tends to be higher in Arizona than in Mexican Pimas, and this may account for a modest proportion of the difference in obesity risk. A marked difference in frequency is observed at HLA, which is involved in type 1 diabetes and other immune diseases; additional studies of the HLA locus are currently underway.